In this thesis undertaking, a comparative analysis between parametric quantities from an bing experimental informations of SCAL measurings and pore web flow theoretical accounts is carried out. The SCAL parametric quantities include comparative permeableness and capillary force per unit area. The experimental dataset was obtained from a big figure of sandstone Fieldss consisting capillary force per unit area and comparative permeableness informations for drainage and imbibition. The curves from different procedures were obtained from suiting theoretical accounts such as: Corey or power jurisprudence ( comparative permeableness ) , Brooks & A ; Corey ( drainage capillary force per unit area ) and Skjaeveland ( primary imbibition capillary force per unit area ) . In add-on, correlativities were established between curve parametric quantities and permeability/porosity and wettability for the complete subset of sandstone stones. In the same manner, a set of statistical practical stones was created utilizing the procedure based Reconstruction technique. The Reconstruction starts from the grain size distribution of certain sedimentary stone type such as Fontainebleau, Bentheim, Berea and a customized North Sea reservoir in order to imitate the natural procedures of organizing sedimentary stones ; i.e. deposit, compression and diagenesis. This technique allowed the creative activity of a sum of 34 sandstone practical stones with different petrophysical belongingss. Subsequently, the pore web theoretical accounts were extracted and a sum of 3210 pore web flow theoretical accounts were obtained with changing oil-wet pore fraction, contact angles ( making 3 different conditions ) and administering oil-wet elements based on pore size: uncorrelated with pore size ( random ) , preferred large-pores and preferred small-pores. The information points obtained from the pore web flow theoretical accounts were fitted by implementing the Marquard-Levenberg and Trust Region algorithm curve suiting techniques utilizing the same theoretical account curves explained before ( Corey, Brooks & A ; Corey and Skjaeveland ) . In a undermentioned measure suiting parametric quantities were plotted against permeability/porosity and wettability index giving a 2nd set of correlativities. Both sets of informations were compared ensuing in understandings in some instances and dissensions in others. It was found that for primary drainage all parametric quantities agreed good ( Pe, Sir ) but the curve form factor a demoing lower tendency values compared with the experimental dataset. The ground might be associated with either the absence of kineticss effects in the pore web theoretical accounts or the unsimilarities in the grain size distribution. Parameters for primary imbibition showed some understandings every bit good as dissensions. In footings of capillary force per unit area parametric quantities, the oil curve form factor ao showed a comparatively good correlativity with regard to experimental informations and proposing that the experimental nucleus sample should be classified as intermediate-wet. In add-on, the H2O curve form factor aw pore web values resulted in slightly higher than experimental informations. It is of import to observe that the experimental value of aw was fixed at 0.2 accounting for non dependable experimental values at low impregnations. With respect to the entry force per unit areas, cw showed comparative similar values as to experimental informations whereas carbon monoxide was somewhat lower. Furthermore, the pore web flow theoretical accounts wettability consequences were compared with the experimental informations utilizing different contact angle ranges, changing the oil-wet pore fraction and altering the distributing of oil-wet elements based on pore size. No clear correlativity was found for mixed-wet and fractionally-wet wettability categories with theoretical tendencies. However, some variables did act upon the wettability of the pore web flow theoretical accounts. Both comparative permeableness curves, form and values showed some sensitiveness when analyzed with different pore-scale supplanting mechanism, oil-wet pore fraction, contact angle and pore size distribution. It is of import to take the possible being of mistakes into history for both, experimental dataset, every bit good as the curve fitting process and some failings related to the pore-scale physical premises in pore web mold.
Presents due to the necessity of understanding how the hydrocarbon recovery could be increased as a effect of the world-wide depletion of conventional oil reservoirs, several elaborate surveies of the reservoirs are carried out.
Normally to obtain a stone sample from a reservoir, a nucleus must be cut from it. Subsequently, the nucleus must be shipped to research lab installations where a sample must be collected. Laboratory experiments performed on reservoir nucleuss have been the chief beginning of informations, nevertheless, these nucleuss have to go through through a long and boring procedure before any information can be acquired. These procedures include cleansing, criterion and particular nucleus analysis, ageing in order to find petrophysical belongingss such as porousness, seawater permeableness, capillary force per unit area, comparative permeableness and others. The most appropriate handling and readying processs vary with the ultimate aim of the nucleus analysis.
From a common nucleus analysis, belongingss such as: porousness, permeableness in different waies, residuary oil impregnation and connate H2O impregnation can be obtained. In contrast, for particular nucleus analysis samples are handled with particular attention to reconstruct the reservoir wettability conditions in order to use measurings. The informations obtained could be: capillary force per unit area, comparative permeableness, wettability and residuary oil impregnation.
Therefore, a tool that could potentially supply the nexus between the reservoir sample and the indispensable information required without holding to execute research lab experiments is used in this work. Its prognostic capablenesss ( still non proved engineering ) are based on pore web theoretical accounts of digital stone.
Network theoretical accounts were pioneered by Fatt in the 1950s [ 3 ] by administering the pores and pharynxs on a regular planar lattice and consecutive make fulling them in the order of inscribed radius utilizing the Young-Laplace equation. During his research, he was able to formalize the pore web theoretical account with experimental informations and to obtain capillary force per unit area and comparative permeableness curves for drainage as a map of mean impregnation.
At the same clip, the classical infiltration theory was introduced by Broadbent and Hammersley [ 4 ] , patterning the random flow of a fluid through a medium and depicting how elements of the medium independently permit or fail to allow transition. However, farther progresss from Fatt ‘s position did non happen until the early 80s, when Larson et Al. [ 5 ] , looking at pore degree natural philosophies combined the infiltration theory with a stochastic attack to foretell the distribution of the non-wetting stage during two stage flow.
Since so, understanding of the implicit in physical mechanisms and the function of pore-space geometry and topology has been increasing. Lenormand et Al. [ 6 ] , K.K. Mohanty et Al. [ 7 ] used theoretical accounts with rectangular cross-section capillaries to detect and depict the supplanting processes during imbibition. These theoretical accounts are still the foundations for web mold: piston-like supplanting, snap-off and pore-body filling, leting the pore-scale natural philosophies of supplanting to be understood.
Thereafter in the early 90s, two-phase and three-phase web theoretical accounts and the ability to include wettability change were developed by several writers. Kovscek et Al. [ 8 ] , incorporated thin wetting movie forces into a aggregation of capillary tubings theoretical account to depict the geological development of alleged mixed-wettability in reservoir stones. Their theoretical account emphasizes the singular function of thin movies. Blunt [ 9 ] extended their analysis to different fractions of oil-wet pores, contact angles and initial H2O impregnation. Sahimi et Al. [ 10 ] , developed a two stage comparative permeableness theory utilizing stochastic constructs from infiltration theory.
The extension of the three -phase flow web theoretical accounts to existent porous media has been complicated by the trouble of depicting the 3-D pore construction of existent stones.
Therefore, the usage of webs derived from a existent porous medium was foremost introduced by Bryant et Al. [ 11 ] by pull outing their web from a random close wadding of every bit sized domains and doing anticipations of comparative permeableness, capillary force per unit area and electrical conduction. Widening this attack for a wider set of sedimentary stones, Oren et Al. [ 12 ] introduced a procedure based method which mimics the terminal province of the geological procedure that formed the stone. This attack has proved its foretelling capablenesss in several plants [ 13, 14, 15 ] .
Nowadays, many groups of research are dedicated to the country of pore web patterning, such as Imperial College, Heriot-Watt University, Australian National University and Numerical Rocks. Digital stone natural philosophies or pore web mold is an emerging field that may add a great potency for the approaching old ages.
In this work, the experimental dataset used was antecedently generated in-house by Shell International Exploration and Production B.V. , happening correlativities between SCAL parametric quantities and porosity/permeability measurings for drainage and imbibition. The capillary force per unit area was obtained by utilizing the extractor method where a cylindrical sample is foremost saturated with oil and mounted in a extractor in order to whirl in stairss of increasing rate. The centrifugal forces throw oil from the sample, while drawing environing H2O into the sample. The continuance of each velocity measure must be sufficient for production of oil to discontinue. In add-on, comparative permeableness measurings were done utilizing steady province methods. In each experiment, fluids are injected at changeless rate and fractional flow. The force per unit area bead and impregnation are non recorded until they reach stable values. Then the fractional flow is changed and the process is repeated.
The consequences obtained from research lab experiments are stored in a database and the information is analyzed looking at some tendencies or correlativities. Smits and Jing [ 2 ] , investigated comparative permeableness curves utilizing the Corey Model or “ power jurisprudence ” . He linked the six Corey parametric quantities to happen out their several correlativities to water/oil imbibition. The survey included 13 sandstone reservoirs. Cense [ 1 ] suggested a set of correlativities utilizing the consequences of laboratory experiments on water/oil primary drainage and imbibition for capillary force per unit area in sandstone and carbonate reservoirs. From these two plants, the best tantrum for drainage is based on the semi-empirical relationship of Brooks and Corey [ 16 ] parametric quantities:
and ( 1.1 )
where Personal computer is the capillary force per unit area in Pascal, Pe is the entry force per unit area in Pascal, Sr is the decreased impregnation, Sir is the irreducible H2O impregnation and a is the curve form factor.
For imbibition and secondary drainage, a Skjaeveland [ 17 ] tantrum is used to depict the capillary force per unit area curves:
( 1.2 )
and ( 1.3 )
where cw is the H2O entry force per unit area, carbon monoxide is the oil entry force per unit area, aw, ao are the H2O and oil curves shape factor severally. The equation 1.2 consists of two subdivisions. The first subdivision is called H2O subdivision and is defined by parametric quantities cw, aw and Srw whereas the parametric quantities co, ao and Sro define the 2nd subdivision or so called oil subdivision.
During this work, practical stones utilizing a procedure based Reconstruction technique are created and their several pore webs are extracted. Once the pore web theoretical account is created, the flow in the web is modeled to obtain for primary drainage, primary imbibition and secondary drainage, the comparative permeableness and capillary force per unit area curves. A set of suiting parametric quantities that describes old curves are correlated with permeableness, porousness and wettability.
Therefore, the importance of this work relies on an alternate attack capable to place failings and strengths resulted from the informations comparing between the web flow mold and research lab experiments.
The workflow procedure of this work is done utilizing the e-Core package ( see appendix A ) . It is chiefly conformed by three phases: the first one is related to the creative activity of the geological theoretical account, where the digital stone samples are generated utilizing petrophysical parametric quantities. The geological theoretical accounts have to go through through the different typical chief organizing sandstone procedures, such as: deposit, compression and diagenesis.
Second, the pore web is extracted from the geological theoretical account where several nodes and links stand foring the pores and pharynxs of the stone are recreated.
Last, the pore web flow theoretical accounts describes the stone fluid interaction occurred at pore-scale. This is done in order to foretell constituent relationships of comparative permeableness and capillary force per unit area. Once the constituent relationships are acquired, some analytical correlativities are used in order to happen a representative tendency between the pore web flow theoretical accounts informations and research lab informations.